Hasegawa et al. – Supplemental Page 1

Supplemental Figures

Supplemental Figure 1. Projections of orexin neurons are preserved in OX1R-/-;OX2R-/-

mice. Coronal brain sections prepared from OX1R-/-;OX2R-/- and wild-type mice were

stained brown with anti-orexin A antibody and counterstained with hematoxylin.

Orexin-immunoreactive nerve fibers are indicated by arrowheads. The locations of the

magnified images are indicated by boxes in the low-power images. Scale bars, 100 µm.

Hasegawa et al. – Supplemental Page 2

Supplemental Figure 2. Region-specific restoration of orexin receptor expression in

OX1R-/-;OX2R-/- mice. (A) Mean duration and number of NREM sleep episodes. (B) Total

time spent in wakefulness and NREM sleep during the dark phase in OX1R-/-;OX2R-/- mice

with or without the targeted restoration of orexin receptor expression in the indicated nuclei

(and the subtype of orexin receptor) and wild-type (WT) mice in the dark phase. Mean

values of three nights’ recordings are shown. Control (1R-/-;2R-/-) was OX1R-/-;OX2R-/- mice

with AAV-CAG/EGFP injection into the DR or LC, AAV-Pet1/ChR2::EYFP in DR, or

AAV-PRSx8/ChR2::EYFP in LC. *P < 0.05, **P < 0.01 vs. control; #P < 0.05, ##P < 0.01

vs. wild-type mice. Values are mean ± S.E. (n = 16 for control, n = 7 for DR-restored, n = 8

for LC-restored, n = 4 for TMN-restored, n = 4 for PPT-restored, n = 5 for PPT-restored, n

= 4 for wild-type mice).

Hasegawa et al. – Supplemental Page 3

Supplemental Figure 3. Neuronal type-selective restoration of orexin receptor expression

in the DR and LC of OX1R-/-;OX2R-/- mice. (A) Mean duration and number of NREM sleep

episodes. (B) Total time spent in wakefulness and in NREM sleep during the dark phase in

OX1R-/-;OX2R-/- mice with control virus injection (1R-/-;2R-/-: same as in Supplemental

Figure 2), those with neuron type-selective restoration of orexin receptor expression

(DR-5HT-2: DR serotonergic-selective OX2R restoration = OX1R-/-;OX2R-/-+5HT-OX2R

mice; LC-NA-1: LC noradrenergic-selective OX1R restoration =

OX1R-/-;OX2R-/-+NA-OX1R mice), OX2R-/-, and wild-type mice (data from the

Supplemental Figure 2 are included for comparison) in the dark phase. Mean values of

three nights’ recordings are shown. *P < 0.05, **P < 0.01 vs. control; #P < 0.05, ##P <

0.01 vs. wild-type mice.

Hasegawa et al. – Supplemental Page 4

Supplemental Figure 4. The restoration of orexin receptor expression in the regions

surrounding the LC does not ameliorate narcoleptic symptoms. (A, B) Expression of

OX1R::EYFP (A) or OX1R::EGFP (B) in OX1R-/-;OX2R-/-+non-NA-OX1R mice (A), which

have similar numbers of EYFP(+)/TH(-) cells but have few EYFP(+)/TH(+) cells as

Hasegawa et al. – Supplemental Page 5

compared to OX1R-/-;OX2R-/-+NA-OX1R mice, and in OX1R-/-;OX2R-/-+PB-OX1R mice (B),

which received the focal injection of AAV-EF1α/OX1R::EGFP in their PB. (C-H)

Sleep/wakefulness parameters of OX1R-/-;OX2R-/-+non-NA-OX1R and

OX1R-/-;OX2R-/-+PB-OX1R mice during the dark phase. (C) Number of and time spent in

cataplexy-like episodes. (D) Mean REM sleep latency and time spent in REM sleep. (E)

Mean duration and number of wakefulness episodes. (F) Mean duration and number of

NREM sleep episodes. (G) Time spent in wakefulness and NREM sleep. *P < 0.05, **P <

0.01 vs. control; #P < 0.05, ##P < 0.01 vs. wild-type mice. (H) Time-weighted frequency

histograms of wakefulness duration showing the proportion of wakefulness that occurred in

episodes of each length to the total amount of wakefulness in the dark phase. 1R-/-;2R-/-,

control mice as in Supplemental Figure 2 (n = 16); LC-1, OX1R-/-;OX2R-/-+LC-OX1R mice

(n = 8); PB-1, OX1R-/-;OX2R-/-+PB-OX1R mice (n = 6); LC-NA-1,

OX1R-/-;OX2R-/-+NA-OX1R mice (n = 5); non-NA-1, OX1R-/-;OX2R-/-+non-NA-OX1R mice

(n = 6); WT, wild-type mice (n = 4).

Hasegawa et al. – Supplemental Page 6

Supplemental Figure 5. Consolidation of fragmented wakefulness is correlated with the

number of noradrenergic neurons restored with OX1R in the LC. (A) Correlation between

the mean duration of wakefulness episodes and the number of EYFP(+) cells in the LC

noradrenergic neurons, or in the region lateral (including parabrachial nucleus: PB), medial,

or ventral (including subcoeruleus region: SubC) to the LC of OX1R-/-;OX2R-/-+NA-OX1R

mice and OX1R-/-;OX2R-/-+non-NA-OX1R mice. Pearson’s correlation coefficients (r), P

values, and regression lines are also shown. (B) Expression of OX1R::EYFP (orange stain)

and NPS mRNA (blue stain) around the LC of OX1R-/-;OX2R-/-+NA-OX1R mice and

OX1R-/-;OX2R-/-+non-NA-OX1R mice, and correlation between the mean duration of

wakefulness episodes and the number of EYFP(+);NPS(+) cells in these mice.

Representative EYFP(+);NPS(-), EYFP(-);NPS(+), and EYFP(+);NPS(+) cells are indicated

by red, blue, and green arrowheads, respectively.

Hasegawa et al. – Supplemental Page 7

Supplemental Figure 6. Pharmacogenetic activations of LC noradrenergic and DR

serotonergic neurons in orexin/ataxin-3 mice. Expression of hM3Dq and sleep parameters

after the injection of CNO or saline in orexin/ataxin-3 mice with DR serotonergic (A) or

LC noradrenergic (B) neuron-selective expression of hM3Dq. For expression analyses,

coronal brain sections containing the DR or LC were initially hybridized in situ to an

hM3Dq antisense probe (red), then were immunostained with anti-TPH or anti-TH

antibodies. The locations of the magnified images are indicated by white arrowheads in the

low-power images (green). The intense red staining indicated by an arrow (B) was likely to

be non-specific, because similar staining was observed also in sections of mice that had not

received AAV injections. Total time spent in REM sleep, wakefulness, and in NREM sleep,

mean duration and number of NREM sleep episodes within 6 h after saline or CNO

administration at ZT12 are shown. *P < 0.05, **P < 0.01; two-tailed Student’s paired t test.

Values are mean ± S.E. (n = 7 for DR, n = 8 for LC).

Hasegawa et al. – Supplemental Page 8

Supplemental Figure 7. Consolidation of fragmented wakefulness by CNO administration

is correlated with the number of hM3Dq-expressing cells in the LC noradrenergic neurons.

(A) The wake-stabilizing effect of CNO was not observed in orexin/ataxin-3 mice with

off-target hM3Dq expression, which have similar numbers of hM3Dq-expressing cells in

the regions surrounding the LC but have almost no such cells in the LC. Expression of

hM3Dq mRNA (A), time spent in, mean duration and number of each stage within 6 h after

Hasegawa et al. – Supplemental Page 9

saline or CNO administration at ZT12 are shown (B), as well as hourly plots of mean

wakefulness duration and number after administrations (C) and time-weighted frequency

histogram of wakefulness duration (D) within 6 h after saline or CNO administration at

ZT12. *P < 0.05, **P < 0.01; two-tailed Student’s paired t test. Values are mean ± S.E. (n

= 10). The intense red staining indicated by an arrow was likely to be non-specific, because

similar staining was observed also in sections of mice that had not received AAV injections.

(E) Correlation between the mean duration of wakefulness episodes within 6 h after CNO

administration and the number of hM3Dq(+) cells in the LC TH(+) neurons, LC TH(-)

neurons, or in the region lateral (including PB) or medial to the LC of orexin/ataxin-3 mice

with LC noradrenergic neuron-selective (Successfully targeted) or off-target

(Unsuccessfully targeted) hM3Dq expression. hM3Dq(+) cells were not detected in the

SubC area. Pearson’s correlation coefficients (r), P values, and regression lines are also

shown.

Hasegawa et al. – Supplemental Page 10

Supplemental Figure 8. DR serotonergic neurons innervate the nuclei involved in the

regulation of sleep/wakefulness and emotion. Coronal brain sections prepared from

OX1R-/-;OX2R-/- mice with the targeted injection of AAV-Pet1/ChR2::EYFP in the DR,

which were analyzed as a part of control mice for orexin receptor restoration experiments,

were stained brown with anti-GFP antibody or double-stained with anti-GFP antibody (red

fluorescence) and one of the antibodies for neuronal type-specific markers (green

fluorescence): TH for dopaminergic neurons of substantia nigra compacta (SNc), TPH for

DR serotonergic neurons, and ChAT for cholinergic neurons of LDT. EYFP-positive fibers

were observed in many areas, with particularly dense innervations to the central and

basolateral nuclei of the amygdala, SNc, and PPT/LDT. The PPT/LDT contains

REM-active cholinergic neurons implicated in the initiation of REM sleep and REM atonia.

Cataplexy is often triggered by strong emotions, which are controlled by the amygdala.

Thus, DR serotonergic neurons may coordinately control multiple brain regions involved in

the regulation of REM sleep and emotion under the control of orexin neurons, to prevent

cataplexy-like episodes. The locations of magnified images are indicated by arrowheads in

the low-power images. 4v, fourth ventricle; Aq, aqueduct; opt, optic tract. Scale bars, 100

µm.